The Emperor’s New Lesion

February 3, 2017

Debating the existence of a lesion in radiology.

So there you are, plugging away at the cases in your worklist (or, as some from the pre-PACS era might still call it, “stack” or “pile”), and along comes a case in which someone previously reported a lesion. Even if the “reason for exam” doesn’t specifically state the study is to follow up the thing, you know you’d darned well better say something about it…among other reasons, because you’ll be getting a request for an addendum on the matter if you don’t mention the abnormality.

But there’s a problem: You don’t see the damned thing.

The tiny lung nodule, the ditzel in one of the solid abdominal viscera, the pituitary adenoma, the hot (or cold) focus on scintigraphy, whatever was previously mentioned, is completely and utterly failing to reveal itself to your eyes. Nor can you see what the heck the previous rad was talking about on the prior study. It almost makes you wish the prior had been unavailable for comparison, so you could blame its absence for your lack of current detection.

Your mileage may vary, as the expression goes, but when I’m reading a case, if I don’t see something pretty quickly, I’m unlikely to find it with prolonged staring at the screen…and that especially seems to be the case when I “know” that there’s supposed to be something there. It seems as if the time I spend vainly attempting to find the purported abnormality increasingly becomes its own obstacle against my subsequently finding anything.

All the while, what starts off as a quiet murmur in my inner dialogue is growing progressively louder and more insistent: Maybe the lesion isn’t really there, and the previous read was wrong.

Not so easy to say that in a report. Putting aside the thorny matter of disagreeing with another doc on the written record (respect for a professional colleague as well as painting a nice, big medicolegal target sign on yourself), there’s the “Emperor’s new clothes” effect.

That is, if one or more other people were able to see something (and it is known that you know they saw it), how defective do you have to be not to see it?

The effect gets compounded if more than one other person claims to have seen the lesion. For instance, the clinician(s) who got ahold of a “not for diagnostic purposes” CD of the CT or MRI which supposedly showed the abnormality. He doesn’t see the lesion, himself. But will he say so? Maybe not; he might be influenced by confirmation bias (that is, the lesion would fit very well with the clinical scenario). Not to mention that the clinician, himself, is just as fallible a human as the rest of us, and subject to the temptation of the Emperor’s new lesion.

It takes a lot of courage, and confidence in one’s abilities, to stand up and say that the Emperor has no lesion, no matter how many others claim to have seen it. But then, as the attendings used to somewhat condescendingly say to us pittance paid house staff back in the day, “That’s why they pay you the big bucks.”

Incidentally, having been on the receiving end of Emperor lesions more than a few times, I bend over backwards to avoid inflicting them on other folks. Whenever I’m reporting on something tiny or otherwise subtle, I make a point of specifying not only physical features of the thing (size, attenuation/signal/echotexture, location, etc.), but also the specific image number so subsequent docs won’t have to dig through quite as much of a haystack before they find my needle…or declare it to be absent.